HumanoidTTT: Test-Time Capability Reuse for Efficient Humanoid Control

Jingtai Yang; Yining Wu; Yanjun Li; Zeyu Zhang; Hao Tang

cs.ROSep 18, 2026

HumanoidTTT: Test-Time Capability Reuse for Efficient Humanoid Control

Authors: Jingtai Yang, Yining Wu, Yanjun Li, Zeyu Zhang, Hao Tang

Organizations: School of Computer Science, Peking University

Abstract

Recent advances in motion generation and whole-body tracking have enabled humanoid robots to execute increasingly diverse motions, yet the same motion capabilities may be requested repeatedly during continual deployment. Reliable reuse is challenging because intervening motions can change the robot's entry state, making previously successful motions unsafe to replay blindly. Meanwhile, validated capabilities accumulate during deployment, while bounded storage requires deciding which ones are worth retaining. To address these challenges, we present HumanoidTTT, a framework for test-time capability reuse in continual humanoid control. Specifically, we introduce Selective Full-Motion Reuse, which authorizes direct reuse of validated complete motions only from certified applicable entry states, allowing accepted reuse to bypass fresh generation. We further introduce Test-Time Capability Consolidation, which adapts which qualified capabilities persist in a bounded Full-Motion Store using subsequent deployment reuse as feedback. Experiments demonstrate zero unsafe accepts and a 16.4 $\times$ end-to-end speedup over fresh generation, while online consolidation improves avoided generator calls by 13.2 per 200 requests over its frozen counterpart. Overall, HumanoidTTT enables reliable and efficient reuse of validated motion capabilities while adaptively retaining useful capabilities throughout continual deployment. Code: https://github.com/AIGeeksGroup/HumanoidTTT. Website: https://aigeeksgroup.github.io/HumanoidTTT.

Figures & tables

Fig. 2: HumanoidTTT. Entry Applicability controls direct reuse of Validated Full-Motion Capabilities. A miss invokes the Frozen Motion Generator followed by Execution-Based Qualification. When a Qualified Capability reaches a full Store, a Store-conditioned Double-DQN chooses SKIP or replacement of an existing entry and adapts online from subsequent deployment reuse utility.

Fig. 3: Consolidation at a qualified full-Store miss. (a) SKIP retains the current Store without admitting the arrival. (b) REPLACE( $j$ ) admits the new capability and evicts entry $j$ . The Double-DQN scorer selects an action using Store-conditioned features.

Fig. 4: Capacity and request dynamics. (a) Calls avoided versus Store capacity. (b) Online-minus-frozen gains across three request families. Error bars show stream-level 95% bootstrap confidence intervals.

Fig. 5: Full-Motion Capability Reuse across Different Entry States in Continual Deployment. Each row follows a continuous acquire A, acquire B, reuse A sequence. The intervening motion changes the robot state before A is requested again, so reuse begins from a changed valid entry state. When Entry Applicability is satisfied, HumanoidTTT reuses the previously acquired complete motion across these changed entry states.

Fig. 6: Full-motion reuse during real-world continual deployment. Each row shows an A–B–A execution sequence: (a) initial execution of motion A; (b) execution of motion B; and (c) reuse of motion A when requested again. The four sequences cover walking forward, turning, waving, and squatting and standing, illustrating reuse after different intervening motions.

Method	Physical success $\uparrow$	Tracking MPJPE $\downarrow$	Joint-limit violation $\downarrow$
Fresh OMG [ 1 ]	80.50	21.24	5.15
OMG + DiT Cache [ 19 ]	82.00	20.62	5.15
TextOp [ 17 ] (G1 adaptation)	62.00	23.07	26.94
TEXEDO [ 18 ] + HoloMotion [ 3 ] ( $N=32$ )	53.00	43.31	15.96
TEXEDO [ 18 ] + SONIC [ 20 ] ( $N=32$ )	58.50	32.46	1.59
HumanoidTTT (Ours)	83.00	20.94	3.08

TABLE I: Motion execution quality under a common deployment protocol. All methods use the same 200-request stream. Success and joint-limit violation rates are reported as percentages; MPJPE is in millimeters. Success is measured over all requests; joint-limit violation rates use executed requests. Tracking MPJPE includes observed prefixes of failed executions. Safety termination may precede a fall. Bold and underlining denote the best and second-best results, respectively.

Method	Accepted coverage $\uparrow$	Safe accepted $\uparrow$	Unsafe accepts $\downarrow$	OOD false accepts $\downarrow$	Replay success $\uparrow$	Cat. $\downarrow$
Task-ID Replay	200/200	58/200	142	887/887	249/600	1
Global Nearest-Entry	178/200	54/200	124	274/887	232/534	1
Per-Motion Nearest-Entry	172/200	52/200	120	277/887	222/516	1
HumanoidTTT (Ours)	180/200	180/200	0	0/887	540/540	0

TABLE II: Selective full-motion reuse reliability. Safe acceptance is evaluated per endpoint: an endpoint is safe for a method only when all three replays of that method’s accepted path satisfy the fixed criterion. Unsafe accepts are accepted endpoints that fail this criterion. Replay success and catastrophic failures are counted per rollout; Cat. denotes catastrophic rollouts. OOD false accepts are evaluated on 887 entry conditions.

Method	Mean $\downarrow$	P50 $\downarrow$	P95 $\downarrow$
Fresh generator [ 1 ]	5092.975	5057.996	5491.174
Generator + ONNX Runtime [ 16 ]	1148.270	1123.820	1235.144
Generator + DiT Cache [ 19 ]	558.916	543.084	629.073
HumanoidTTT (Ours)	38.409	29.514	63.115

TABLE III: Accepted-reuse preparation latency on confirmed hit contexts. Latency includes Entry Applicability and stored-motion retrieval. Lower is better.

Method	Mean E2E latency $\downarrow$	Average compute $\downarrow$	Peak memory $\downarrow$
Fresh generator [ 1 ]	4415.82	685.97	1.529
Generator + ONNX Runtime [ 16 ]	1047.51	685.97	1.593
Generator + DiT Cache [ 19 ]	523.41	277.44	1.593
TextOp [ 17 ]	1473.26	34.11	1.934
TEXEDO [ 18 ] (Best-of-8)	28595.71	243.74	2.617
HumanoidTTT (Ours)	268.66	75.46	1.595

TABLE IV: End-to-end inference efficiency in FP32 on the fixed 200-request high-recurrence trace. Latency spans request arrival to an executable G1 reference. Lower is better.

Method	Calls avoided $\uparrow$	SD	Avoided/ slot $\uparrow$	Generator calls $\downarrow$
LRU	97.800	22.904	9.780	102.200
LFU	110.467	21.709	11.047	89.533
One-Step Greedy	112.933	21.697	11.293	87.067
Contextual Bandit	109.867	23.130	10.987	90.133
Pretrained RL, Frozen	118.200	20.232	11.820	81.800
HumanoidTTT (Ours)	131.400	16.017	13.140	68.600

TABLE V: Finite-capacity capability consolidation at

K=10

. Results are means over 15 matched 200-request streams; all components except the consolidation policy are held fixed.

Humanoid Robot Control

Humanoid robot control develops methods for coordinating bipedal balance, locomotion, and whole-body interaction with people and objects.

See all 206 papers

TEXEDO : Test Time Scaling for Controller-aware Language-conditioned Humanoid Motion Generation
Jun 22, 2026Jianuo Cao, Yuxin Chen, Yuzhen Song +3Humanoid Robot Control Test-Time Scaling
Nanjing University · University of California, Berkeley · Southern University of Science and Technology
Whole-Body Humanoid Control via Stochastic Motion Tokenization
Oct 27, 2026Michael Drolet, Oleg Arenz, Jan PetersHumanoid Robot Control Humanoid Motion Tracking
Technical University of Darmstadt · German Research Center for AI (DFKI) · Robotics Institute Germany (RIG)
Humanoid-GPT: Scaling Data and Structure for Zero-Shot Motion Tracking
Jun 2, 2026Zekun Qi, Xuchuan Chen, Dairu Liu +10Zero-Shot Learning Transformer
Tsinghua University · Galbot Inc. · Shanghai Jiao Tong University +2

Jun 22, 2026J/K move · Enter open · S save

TEXEDO : Test Time Scaling for Controller-aware Language-conditioned Humanoid Motion Generation

Jianuo Cao, Yuxin Chen, Yuzhen Song, Masayoshi Tomizuka+2

Nanjing University · University of California, Berkeley · Southern University of Science and Technology

Humanoid Robot Control Test-Time Scaling Human Motion Generation Robot Motion Generation

Open paper AskPDF arXiv

Text-conditioned motion generation is a promising interface for programming humanoid robots, yet current generators are often trained on human motion datasets retargeted to robot morphologies. Although such data provides rich semantic and kinematic priors, it fails to capture the nuances of whole-body tracking controllers, including balance, contact dynamics, actuation limits, and controller-specific failure modes. As a result, generated motions can be semantically plausible but difficult or impossible for the robot to execute. We introduce TEXEDO, a test-time scaling framework for humanoid motion generation that improves motion quality without requiring a stronger underlying generator. Given a text prompt, TEXEDO samples multiple candidate motions from a pretrained text-conditioned generator and selects the best motion that is both executable and task-aligned. The reward model combines a dynamic feasibility verifier, distilled from whole-body tracking rollouts to predict physical executability, with a semantic alignment verifier that measures text-motion alignment in a learned co-embedding space. Our pipeline treats dynamic feasibility as a hard constraint and semantic alignment as the selection objective within the feasible set. Through large-scale simulation studies and real-world deployment on a Unitree G1 humanoid robot, we show that TEXEDO consistently improves both tracking fidelity and text alignment. These results demonstrate that grounded verification is an effective path toward deployable language-guided humanoid motion generation. Project website: https://jianuocao.github.io/TEXEDO/

Abstract

Humanoid Robot Control

Explore similar work

TEXEDO : Test Time Scaling for Controller-aware Language-conditioned Humanoid Motion Generation

Whole-Body Humanoid Control via Stochastic Motion Tokenization

Humanoid-GPT: Scaling Data and Structure for Zero-Shot Motion Tracking